Open systems innovation is the design philosophy of the early internet era — succeeding by interconnecting diverse systems rather than owning every layer — under which value spreads broadly across an ecosystem instead of accruing to a single integrator.
Before the internet, computing was vertically integrated: computer systems, telephony, and video were separate stacks that didn't interact, each owned end-to-end by a large company that captured its benefit. The internet, Unix, and the web inverted that model. In the account 📝Judy Estrin gives on "📝Don't Be Evil" (📝Everything They Know, Ep. 1, cited with permission of host 📝Ari Andersen), the era's founding philosophy "was about interconnecting diversity as opposed to everything having to be the same and big and powerful — that the way you succeeded was by connecting with others. You can think of it as collaborative as opposed to authoritarian."
The economics are the heart of the concept: "When you're vertically integrated, the value accrues to the integrator. When you have open systems innovation, the value is spread more broadly." In a 2011 presentation Estrin placed social, mobile, and cloud at a fork — open or vertically integrated? — and marked it with a question mark. She looks back "with some sadness": the industry reverted to vertical integration, and scale technologies compounded the capture, since network effects and data-hungry AI reward whoever already holds the position.
The same fork now stands before AI and personal-knowledge infrastructure, which is why the concept matters beyond history: any layer that gets better with scale will drift toward integration unless its architecture deliberately spreads the value outward.
We hold this as a design constraint for 📝MythOS: an external brain must be the interconnector — import freely, mirror both ways, value accruing to the library's owner — or it becomes the next integrator.